一种改进的自适应和声搜索优化算法

An improved self-adaptive harmony search optimization algorithm

在传统和声搜索优化算法的基础上,提出一种自适应双子和声搜索优化算法。通过建立主辅两个和声库正反双向进行迭代搜索,并对和声算法中的音调调整概率和音调微调带宽两个重要参数进行自适应调整,提高了算法的动态适应性以及局部搜索和全局搜索的协调能力。通过构造两组搜索方向各异、相互协同的主、辅和声,充分利用了搜索域内的隐含信息,扩展搜索范围,从而实现了全局最优。在试验中分别对3个复杂函数进行测试,结果表明该算法具有较好的全局搜索能力和收敛速度,在一定程度上提高了最优值的搜索能力,达到了预期效果。

Based on the traditional harmony search optimization algorithm,a search optimization algorithm of self-adaptive gemini harmonies was proposed. By establishing two primary harmony and secondary harmony libraries,iterative searching along positive and negative directions,and adaptively adjusting two important parameters of pitch adjusting rate and band width of optimization algorithm,the algorithms dynamic adaptive ability and the coordination ability between local search and global search was improved. Two primary harmony and secondary harmony were constructed that were in different directions and cooperative,which made full use of the hidden information in the search domain,enhanced the searching area,and realized the global optimal. Three complex functions was tested in experiment,the results showthat the algorithm had better global search ability and convergence rate compared with the original algorithm,the search ability of the optimal value was improved to some extent,which achieved the anticipated effects.